The association between social support through contacts with Accredited Social Health Activists (ASHAs) and antenatal anxiety among women in Mysore, India: a cross-sectional study

Abstract

Purpose:

In India, antenatal anxiety prevalence estimates range from 6% to 48%. Social support is strongly associated with mental wellbeing, yet most studies have examined the impact of support from partners and family members rather than peers, community members, or health care providers. This study explores the supportive role of Accredited Social Health Activists (ASHA) contacts for antenatal anxiety.

Methods:

Data were analyzed from the Saving Children, Improving Lives project, a quasi-experimental study conducted among rural, pregnant women in India. Regression models were used to estimate adjusted prevalence ratios (aPRs) and 95% confidence intervals for the relationship frequency of ASHA contacts and antenatal anxiety. Antenatal anxiety was measured using a subscale of the Edinburgh Postnatal Depression Scale.

Results:

The sample consisted of 480 pregnant women. Reported antenatal anxiety prevalence was 27% (95% CI: 23%, 31%). Participants who were more frequently visited by ASHAs at home (aPR: 0.90; 95% CI: 0.76, 0.98) and more frequently accompanied by ASHAs to their antenatal care visits (aPR: 0.86, 95% CI: 0.78, 0.95) were less likely to report antenatal anxiety. ASHA home visits were protective for the most vulnerable women (primigravida and those experiencing domestic violence) and ASHA accompaniment to antenatal care visits was equally protective for all women.

Conclusions:

ASHAs are valued for their contribution towards maternal health education and linking women of reproductive age to healthcare services. Our findings additionally suggest the important role ASHAs play in providing social support to pregnant women, particularly those who are most vulnerable to experiencing antenatal anxiety.

INTRODUCTION

Pregnancy is a time of immense physiological and psychosocial change for women, increasing their vulnerability for the onset or relapse of mental health issues.¹ Some women may experience their first depressive or anxious episode during the antenatal period, while those with a history of depression or anxiety are at increased risk for recurrence, continuation, or intensification.^2,3 Depression and anxiety during pregnancy have been linked with a number of adverse outcomes for both women and their children. Among women, antenatal depression and anxiety been associated with maternal malnutrition, increased substance abuse, and low uptake of antenatal care.^4–7 Children of women with antenatal depression and anxiety are at a higher risk of premature birth, low birthweight, impaired cognitive growth, and development of depression in later life.^8,9 Antenatal depression and anxiety have also been recognized as strong predictors of postnatal depression, one the causes of infant failure to thrive.^1,10–12 Taken together, this evidence points to the importance of developing preventive and therapeutic mental health interventions for women during pregnancy.

The importance of antenatal mental health interventions may have even greater relevance in low- and middle-income countries (LMICs) compared to high-income countries (HICs) given the gap between the burden of mental health issues and available evidence-based services in LMICs.^13–15 Recent systematic reviews and meta-analyses have found that the prevalence of antenatal depression and anxiety can range from 7% to 15% in HICs and from 19% to 25% in LMICs.^13,14,16 While one out of five persons with depression receive care in HICs, only one out of 27 receive any care in LMICs, often due to a lack of funding, human resources, medication, and stigma relating to mental health care.^15,17–21 This is particularly the case in India where antenatal depression and anxiety prevalence estimates range from 6% to 48% and a recent national mental health survey reported a treatment gap of 85% across the country for common mental disorders.^22–29 Among studies from India, few have specifically focused on antenatal anxiety and its associated risk factors. Between the two existing studies, risk factors for antenatal anxiety include economic deprivation, low education, younger age, primigravida, male infant gender preference, intimate partner violence (IPV), domestic violence (DV).^30,31

No studies in India have examined the association between antenatal anxiety and social support, despite considerable evidence in other contexts of its protective value.^7,32–34 Social support is the provision of emotional, instrumental, or informational assistance leading one to believe that they are cared for, loved, esteemed, and a member of a network of mutual obligations.³⁵ Social support may directly protect against anxiety or function by buffering the impact of stressful life circumstances that lead to anxiety.^35–38 For example, experience of IPV is a risk factor for antenatal anxiety and its effect has shown to be buffered by social support.^7,36,39 Studies examining antenatal anxiety have often only measured husband or husband’s family based social support rather than social support from peers, community members, or health care providers.^7,32,33 In India, Accredited Social Health Activists (ASHA) are a cadre of community health volunteers who receive performance-based incentives for promoting universal immunization, referral and escort services for reproductive and child health and other healthcare programs in their villages.⁴⁰ They have the potential to provide a unique source of social support given that they are often trusted women who are invested in improving the wellbeing of their fellow community members, and serve as a link to the local health system. Understanding the role of non-family member social support in this context, specifically from ASHAs, has the potential to inform the development of future evidence-based interventions for the prevention and treatment of antenatal anxiety.

The first aim of this analysis was to examine whether the frequency of contact with ASHAs, serving as a proxy for social support, was associated with antenatal anxiety among a sample of pregnant women in rural India. The second aim of this analysis was to assess the extent to which frequency of contact with ASHAs was protective for women most vulnerable to antenatal anxiety by investigating whether IPV, DV, and gravidity were moderators of the association between ASHA contact and antenatal anxiety.

METHODS

Parent Study Context and Design

Data for this analysis come from the Saving Children, Improving Lives (SCIL) project, a quasi-experimental study conducted among rural, pregnant women in Mysore district of Karnataka, India from 2011 to 2014. Mysore district has a mostly rural (58%) population of three million, half of whom are female.⁴¹ The SCIL project used a combination of community education efforts, mobile medical clinics, and social mobilization via self-help groups to increase uptake of antenatal care and HIV testing pregnant women. Additional details on the design and results of the SCIL project have been published previously.^42,43

Parent Study Recruitment, Data Collection, and Ethics Approval

The SCIL project included 40 villages out of 209 total villages in Mysore district. Villiages were randomly selected using probability proportional to size sampling from a sampling frame defined by the following criteria: located 10 kilometers outside of Mysore city boundaries, a population size between 1,500–3000 individuals, and no existing public medical facility. Community education meetings were first conducted in each selected village to emphasize the importance of antenatal care and HIV testing and to inform members about the SCIL project. Mobile clinics then visited each village. Pregnant women who came to access health services at the mobile clinic were informed about the study, consented, and enrolled, if interested and eligible. Eligibility criteria included being currently pregnant, older than 18 years of age, and residing in a study village for more than six months.

Behavioral survey data and clinical data were collected from participants across three waves of the SCIL project. This analysis used behavioral survey data from the first wave of data collection, spanning April 2011 to March 2012. Surveys were administered by trained interviewers in Kannada, the local language, using structured questionnaires. The behavioral survey contained questions about demographics, socio-economic status, knowledge and attitudes about antenatal care, past and current care-seeking behaviors, interactions with healthcare workers, and psychosocial symptoms. Following the mobile clinic visit, survey data were entered into a Microsoft Access database by trained data entry staff (Microsoft Corporation, Redmond, WA).

The SCIL project received approval from the Institutional Review Board of the Public Health Research Institute of India (protocol number 2011–03-26–10) and Florida International University. Voluntary written informed consent was obtained from all participating women.

Measures

Outcome:

Antenatal anxiety symptoms were evaluated using a three-item sub-scale of the Edinburgh Postnatal Depression Scale (EPDS) known as EPDS-3A.⁴⁴ The original EPDS is a 10-item instrument designed to identify symptoms of depression among women who are pregnant or postpartum.^45–47 EPDS-10, which includes EPDS-3A items, has been previously translated, back-translated, and validated in Kannada among antenatal women in rural Karnataka.⁴⁸ EPDS-3A has been found to be a valid tool for screening for anxiety in pregnant women and reliably able to distinguish between anxiety and depression symptoms. The three items in the EPDS-3A each included a choice of four answers that reflected how the participant was feeling over the past seven days. Individual item scores are recorded as zero, one, two, or three according to symptom severity. Total scores ranged from zero to nine and a cutoff score of six was utilized for detecting antenatal anxiety.⁴⁹

Exposures:

Frequency of contact with ASHAs served as a proxy measure for social support in this analysis using two categorical items. Item one measured the frequency at which ASHAs visited the participant at home. Item two measured the frequency at which ASHAs accompanied the participant to antenatal care visits. Responses to both items included never, rarely (every three months), occasionally (every two months), and regularly (once a month).

Moderators:

Our moderators of interest were IPV, DV, and gravidity. Husband perpetrated IPV and husband’s family perpetrated DV were measured using the three-item Abuse Assessment Screen.^50,51 Participants were asked whether they had ever been “kicked, hit, slapped, or otherwise physically hurt”, “forced into sexual activity against your wish”, or “humiliated or emotionally abused” by their husband or by their husband’s family. Ever experience of husband perpetrated IPV was recorded as positive if the participant answered yes to any of the three husband referent items and null if the participant answered no to all of the three husband referent items. Ever experience of husband’s family perpetrated DV was recorded as positive if the participant answered yes to any of the three husband’s family referent items and null if the participant answered no to all of the three husband’s family referent items. Gravidity was a dichotomous item measuring whether the participant was primigravida versus having been multigravida (previously pregnant).

Covariates:

Our covariates of interest included age (in continuous single years), primary school completion (standard 8), monthly income (in continuous Indian rupees), trimester, and number of living children. These variables have been shown to be associated with antenatal mental health among pregnant women in India.⁵²

Data Analysis

We first calculated the number and proportion of participants who had each sociodemographic characteristic as well as the number and proportion who experienced pregnancy, childbirth, IPV, and DV. We then used chi-square tests to compare characteristics and experiences across participants with antenatal anxiety symptoms and participants without antenatal anxiety symptoms. Where cells were less than five, Fisher’s Exact tests were used (Table 1).

Table 1.

Participant Characteristics and Experiences by Antenatal Anxiety (N = 480)

	All Participants* (N = 480)		Anxious (N = 129)		Not Anxious (N = 351)		Chi-Square
	N	(%)	N	(%)	N	(%)	p-value
Demographics
Age (years)
18–22	336	70%	81	63%	255	73%
23–42	144	30%	48	37%	96	27%	0.037
Education Attained
Primary School Complete	274	57%	64	50%	210	60%
Primary School Incomplete	206	43%	65	50%	141	40%	0.045
Monthly Income
< 5000 INR	267	56%	80	62%	187	53%
5000–10,000 INR	163	34%	36	28%	127	36%
> 10,000 INR	50	10%	13	10%	37	11%	0.048
Pregnancy and Children
Current Trimester
First (1–13 Weeks)	58	12%	8	6%	50	14%
Second (14–26 Weeks)	248	52%	74	57%	174	50%
Third (26–42+ Weeks)	174	36%	47	36%	127	36%	0.046
Gravida
Primigravida	311	65%	94	73%	217	62%
Multigravida	169	35%	35	27%	134	38%	0.025
Number of Living Children
0	260	54%	75	58%	185	53%
>1	220	46%	54	42%	166	47%	0.290
Violence Experience
Husband
Physical	61	13%	28	22%	33	9%	p < 0.001
Sexual	54	11%	30	23%	24	7%	p < 0.001
Emotional	62	13%	29	22%	33	9%	p < 0.001
Husband’s Family
Physical	12	3%	10	8%	2	1%	p < 0.001
Sexual	1	0%	1	1%	0	0%	p < 0.001
Emotional	70	15%	34	26%	36	10%	p < 0.001

^*Columns do not add to total due to missing data

Second, we used a generalized linear model with a log link and binomial distribution to estimate prevalence ratios and corresponding 95% confidence intervals for the relationship between ASHA support and antenatal anxiety. We ran unadjusted and covariate adjusted models. The covariate adjusted models included age in years, primary school completion, monthly income, trimester, gravidity, and ever experience of IPV or DV (husband and husband’s family perpetrated). To explore our effect modification hypothesis, we refit our covariate adjusted model with interaction terms between the two types of ASHA support and gravida as well as the two types of ASHA support and having ever experienced IPV or DV (husband and husband’s family perpetrated). We then performed Wald-type F-tests to assess the joint statistical significance of the interaction terms. All analyses were conducted using StataSE, version 14.2 (College Station, TX).

RESULTS

Population Characteristics

The sample consisted of 480 pregnant women who were predominantly primigravida (65%) and were in their second trimester (52%). Median age was 21 years (interquartile range 18–23 years). The majority of participants had completed primary education (57%) and did not have living children (55%). Median monthly income 5000 INR (interquartile range 3000–8000 INR) and most participants had a monthly income less than 5000 INR per month (56%). Participants reported ever experiencing husband perpetrated physical (13%), sexual (11%), and emotional (13%) violence, as well as ever experiencing husband’s family perpetrated physical (3%), sexual (0.01%), and emotional (15%) violence (Table 1).

Population Characteristics and Antenatal Anxiety Prevalence

Prevalence of reported antenatal anxiety was 26.8% (95% CI: 0.23, 0.31). Participants with antenatal anxiety differed from participants without antenatal anxiety across most sociodemographic characteristics, experiences of pregnancy, and experiences of IPV. Compared to participants without anxiety, participants with anxiety were younger (median age 19 vs. 21, p = 0.032), a lower proportion had completed primary school (50% vs. 60%, p = 0.045), and higher proportion lived in households with monthly incomes less than 5000 INR (62% vs. 53%, p = 0.048).

ASHA Support and Antenatal Anxiety

Both types of ASHA support were negatively associated with prevalance of antenatal anxiety. Participants who were more frequently visited by ASHAs at home (adjusted prevalence ratio (aPR): 0.90; 95% CI: 0.76, 0.98) and participants who were more frequently accompanied by ASHAs to their ANC visits (aPR: 0.86, 95% CI: 0.78, 0.95) were less likely to have antenatal anxiety (Table 2 and Table 3).

Table 2.

Association Between ASHA Home Visits and Antenatal Anxiety (N = 480)

	Model 1				Model 2
	β	95% Cl	aRR	95% Cl	β	95% Cl	aRR	95% Cl
Main Effects
ASHA Home Visits	−0.16*	(−0.35,−0.02)	0.90*	(0.76,0.99)	−0.17*	(−0.53,−0.02)	0.90*	(0.73,0.99)
Age	0.04	(−0.03,0.11)	1.03	(0.97,1.10)	0.04	(−0.03,0.11)	1.04	(0.97,1.10)
Primary Education Complete	−0.02	(−0.08,0.04)	0.99	(0.95,1.03)	−0.01	(−0.07,0.05)	0.99	(0.95,1.03)
Monthly Income (INR)	−0.11	(−0.44,0.22)	0.93	(0.74,1.16)	−0.12	(−0.46,0.21)	0.92	(0.74,1.15)
Trimester	0.19	(−0.14,0.51)	1.15	(0.89,1.49)	0.18	(−0.15,0.51)	1.14	(0.88,1.46)
Primigravida	0.25*	(0.02,0.99)	1.19*	(1.03,1.67)	−0.05	(−1.12,1.02)	0.97	(0.48,1.96)
Experienced Any DV (Husband)	0.69*	(0.10,1.27)	1.58*	(1.08,2.30)	1.20	(−0.34,2.75)	2.13	(0.92,4.90)
Experienced Any DV (Husband’s Family)	0.73*	(0.10,1.36)	1.62*	(1.09,2.40)	2.07*	(0.35,3.81)	3.05*	(1.70,5.49)
Interactions
ASHA Home Visits X Primigravida					−0.13*	(−0.29,−0.06)	0.91*	(0.74,0.98)
ASHA Home Visits X DV (Husband)					−0.55*	(−1.20,−0.11)	0.73*	(0.54,0.99)
ASHA Home Visits X DV (Husband’s Family)					−0.19*	(−0.78,−0.04)	0.89*	(0.62,0.99)

^*p-value < 0.05

Table 3.

Association Between ASHA ANC Visits and Antenatal Anxiety (N = 480)

	Model 1				Model 2
	β	95% Cl	aRR	95% Cl	β	95% Cl	aRR	95% Cl
Main Effects
ASHA ANC Visits	−0.23*	(−0.40,−0.06)	0.86*	(0.78,0.95)	−0.24*	(−0.57,−0.09)	0.89*	(0.70,0.99)
Age	0.04	(−0.03,0.11)	1.04	(0.97,1.11)	0.04	(−0.03,0.11)	1.03	(0.97,1.10)
Primary Education Complete	−0.02	(−0.08,0.04)	0.99	(0.95,1.03)	−0.02	(−0.08,0.04)	0.99	(0.95,1.02)
Monthly Income (INR)	−0.13	(−0.46,0.20)	0.92	(0.73,1.15)	−0.12	(−0.46,0.21)	0.92	(0.74,1.15)
Trimester	0.15	(−0.18,0.48)	1.12	(0.87,1.44)	0.13	(−0.20,0.47)	1.10	(0.86,1.40)
Primigravida	0.23*	(−0.76,−0.07)	1.17*	(1.08,1.93)	0.79*	(0.08,1.51)	1.72*	(1.04,2.83)
Experienced Any DV (Husband)	0.77*	(0.17,1.37)	1.6*	(1.14,2.41)	1.25*	(0.27,2.23)	2.15*	(1.28,3.6
Experienced Any DV (Husband’s Family)	0.74*	(0.09,1.38)	1.61*	(1.09,2.39)	0.73*	(0.18,1.76)	1.59*	(1.01,2.94)
Interactions
ASHA ANC Visits X Primigravida					0.43	−0.79,0.07)	0.7	(0.77,1.01)
ASHA ANC Visits X DV (Husband)					−0.30	−0.78,0.17)	0.82	(0.62,1.10)
ASHA ANC Visits X DV (Husband’s Family)					−0.02	(−0.48,0.52)	0.96	(0.73,1.41)

^*p-value < 0.05

ASHA Support, Antenatal Anxiety, and Gravidity

Primigravida were more likely to have antenatal anxiety in the ASHA home visit adjusted model (aRR: 1.19, 95% CI: 1.03, 1.67) as well as the ASHA ANC visit adjusted model (aRR: 1.17, 95% CI: 1.08, 1.93). Gravidity was a modifier of ASHA home visits (Interaction Beta: −0.13, 95% CI: −0.29, −0.06) but not for ASHA ANC visits (Interaction Beta: −0.43, 95% CI: −0.79, 0.07) (Table 2 and Table 3). Our effect modification results suggest that ASHA home visits were only protective for primigravida and ASHA ANC visits were protective for all women, regardless of gravidity (Figure 1a-1b).

Figure 1a-1b.

ASHA Social Support and Antenatal Anxiety by Gravida

ASHA Support and Antenatal Anxiety Stratified by Husband Perpetrated IPV

Participants who had experienced any husband perpetrated IPV were more likely to have antenatal anxiety in our ASHA home visit adjusted model (aRR: 1.58, 95% CI: 1.08, 2.30) as well as our ASHA ANC visit adjusted model (aRR: 1.66, 95% CI: 1.14, 2.41). Husband perpetrated IPV was a modifier of ASHA home visits (Interaction Beta: −0.55, 95% CI: −1.20, −0.11) but not for ASHA ANC visits (Interaction Beta: −0.30, 95% CI: −0.78, 0.17) (Table 2 and Table 3). Our effect modification results suggest that ASHA home visits were only protective for participants who had experienced husband perpetrated IPV and ASHA ANC visits were protective for all women, regardless of husband perpetrated IPV experience (Figure 2a-2b).

Figure 2a-2b.

ASHA Social Support and Antenatal Anxiety by Husband Perpetrated Domestic Violence

ASHA Support and Antenatal Anxiety Stratified by Husband’s Family Perpetrated DV

Participants who had experienced any husband’s family perpetrated DV were more likely to have antenatal anxiety in our ASHA home visit adjusted model (aRR: 1.62, 95% CI: 1.09, 2.40) as well as our ASHA ANC visit adjusted model (aRR: 1.61, 95% CI: 1.09, 2.39). Husband’s family perpetrated DV was a modifier of ASHA home visits (Interaction Beta: −0.19, 95% CI: −0.78, −0.04) but not for ASHA ANC visits (Interaction Beta: −0.02, 95% CI: −0.48, 0.52) (Table 2 and Table 3). Our effect modification results suggest that ASHA home visits were only protective for participants who have experienced husband’s family perpetrated DV and ASHA ANC visits were protective for all women, regardless of husband’s family perpetrated DV experience (Figure 3a-3b).

Figure 3a-3b.

ASHA Social Support and Antenatal Anxiety by Husband’s Family Perpetrated Domestic Violence

DISCUSSION

To our knowledge, this study is one of the first to examine the association between ASHA contacts and antenatal mental health outcomes in India. More than a quarter of all study participants reported antenatal anxiety. Participants who were more frequently visited by ASHAs at home and participants who were more frequently accompanied by ASHAs to their antenatal care visits were less likely to report antenatal anxiety. In addition, ASHA home visits were protective for the most vulnerable women (primigravida and those experiencing husband or husband’s family perpetrated intimate partner violence) and ASHA accompaniment to antenatal care visits was equally protective for all women. Taken together, our results suggest the potentially important role ASHAs play in providing contact-based social support to pregnant women, particularly those who are most vulnerable to experiencing antenatal anxiety.

The linkage between poor partner and family social support and antenatal depression has been well documented in India but non-family member social support is largely unexamined.^23,54–57 Non-family member social support might be especially salient given the marital and household structure across many parts of the country. Women often leave their parental homes after marriage and join the multigenerational household of their spouse, exposing them to the norms and fertility preferences of their new family members. Pressure to conform to in-law expectations and to produce male children also has been found to be isolating and associated with antenatal anxiety and depression.^23,58–61 Given this context, identifying non-family members who provide social support and facilitating their efforts could be equally as important as improving partner and family social support for pregnant women. Our findings add to the existing literature by examining ASHA social support and suggest its potential for protection against poor mental health outcomes.

Psychosocial interventions delivered by non-specialists have been found to be beneficial for common perinatal mental disorders in low and middle income countries.^62,63 Many of these interventions have included a component of peer support provision by trained lay health workers or “mentor mothers”. ^62,64,65 Lay peer support is built on the well documented effects of social support on health, as well as the recognition that people experiencing disease or other health challenges have a great deal to offer each other.^38,66–69 In addition to knowledge and experience of a certain health concern or stressor, peer supporters often share a similar demographic background and community network with support recipients, enhancing their ability to provide socially relevant care and assistance.⁷⁰ ASHAs are uniquely positioned to provide peer support given that they are trusted women who serve as a link between communities and health systems. Due to the observational nature of our study, we did not collect data on the characteristics of the ASHAs or the nature of their interactions with participants during home visits or accompaniment to ANC visits. Future research that documents shared characteristics and the type of support provided is essential for developing effective peer support components for antenatal anxiety and depression interventions.

Our antenatal anxiety prevalence results slightly differed from prior cross-sectional studies of antenatal anxiety in India. Beyond differences in population, all four studies used different tools to measure antenatal anxiety: the Pregnancy Related Thoughts Scale, the State Trait Anxiety Inventory, the Hospital Anxiety and Depression Scale, and the Pregnancy Related Anxiety Scale.^26,31,71,72 These tools differ from the EPDS-3A as they include more than three items, physical symptoms, and suggest cut-points to scale the severity of antenatal anxiety. Although the EPDS-3A has been validated among pregnant women in southern India, future work that explores the differences among antenatal anxiety scales in this population is necessary.⁴⁸

Our study has a few limitations. First, with our cross-sectional data, we report the prevalence of reported antenatal anxiety, but cannot make causal inferences about how any of the considered variables affect incidence and duration of reported antenatal anxiety. Our data do not show whether reported antenatal anxiety was incident during pregnancy or preceded the pregnancy. Second, we utilized contact with ASHAs as a proxy measure for social support which is a complex construct. Third, the EPDS, and the EPDS-3A subscale, are screening tools and do not diagnose antenatal anxiety or depression. While our cut-point of six was based on the validation study and demonstrated reasonable sensitivity and specificity, it is possible that some participants were falsely categorized as screening positive or negative for reported antenatal anxiety.⁴⁴ The cross-sectional nature of the study imposed a fourth limitation. Anxiety symptoms may vary over time during a woman’s pregnancy but our assessment was restricted to a single point in time. Fifth, data were based on self-report which can be unreliable due to social desirability issues, perception, or memory challenges.Finally, our social support measures were derived from single items on the behavioral survey. Multidimensional scales may better capture these variables of interest more accurately.

CONCLUSION

ASHAs are valued for their contribution towards maternal health education and linking women of reproductive age, and their children, to healthcare services. Our findings additionally suggest the important role ASHAs play in providing social support to pregnant women, particularly those who are most vulnerable to experiencing antenatal anxiety.